Adaptive unscented Kalman filter for neuronal state and parameter estimation.

Data assimilation techniques for state and parameter estimation are frequently applied in the context of computational neuroscience. In this work, we show how an adaptive variant of the unscented Kalman filter (UKF) performs on the tracking of a conductance-based neuron model. Unlike standard recursive filter implementations, the robust adaptive unscented Kalman filter (RAUKF) jointly estimates the states and parameters of the neuronal model while adjusting noise covariance matrices online based on innovation and residual information. We benchmark the adaptive filter's performance against existing nonlinear Kalman filters and explore the sensitivity of the filter parameters to the system being modelled. To evaluate the robustness of the proposed solution, we simulate practical settings that challenge tracking performance, such as a model mismatch and measurement faults. Compared to standard variants of the Kalman filter the adaptive variant implemented here is more accurate and robust to faults.

Modeling Reveals Human-Rodent Differences in H-Current Kinetics Influencing Resonance in Cortical Layer 5 Neurons.

While our understanding of human neurons is often inferred from rodent data, inter-species differences between neurons can be captured by building cellular models specifically from human data. This includes understanding differences at the level of ion channels and their implications for human brain function. Thus, we here present a full spiking, biophysically detailed multi-compartment model of a human layer 5 (L5) cortical pyramidal cell. Model development was primarily based on morphological and electrophysiological data from the same human L5 neuron, avoiding confounds of experimental variability. Focus was placed on describing the behavior of the hyperpolarization-activated cation (h-) channel, given increasing interest in this channel due to its role in pacemaking and differentiating cell types. We ensured that the model exhibited post-inhibitory rebound spiking considering its relationship with the h-current, along with other general spiking characteristics. The model was validated against data not used in its development, which highlighted distinctly slower kinetics of the human h-current relative to the rodent setting. We linked the lack of subthreshold resonance observed in human L5 neurons to these human-specific h-current kinetics. This work shows that it is possible and necessary to build human-specific biophysical neuron models in order to understand human brain dynamics.

Deciphering how interneuron specific 3 cells control oriens lacunosum-moleculare cells to contribute to circuit function.

The wide diversity of inhibitory cells across the brain makes them suitable to contribute to network dynamics in specialized fashions. However, the contributions of a particular inhibitory cell type in a behaving animal are challenging to untangle as one needs to both record cellular activities and identify the cell type being recorded. Thus, using computational modeling and theory to predict and hypothesize cell-specific contributions is desirable. Here, we examine potential contributions of interneuron-specific 3 (I-S3) cells-an inhibitory interneuron found in CA1 hippocampus that only targets other inhibitory interneurons-during simulated θ rhythms. We use previously developed multicompartment models of oriens lacunosum-moleculare (OLM) cells, the main target of I-S3 cells, and explore how I-S3 cell inputs during in vitro and in vivo scenarios contribute to θ. We find that I-S3 cells suppress OLM cell spiking, rather than engender its spiking via postinhibitory rebound mechanisms, and contribute to θ frequency spike resonance during simulated in vivo scenarios. To elicit recruitment similar to in vitro experiments, inclusion of disinhibited pyramidal cell inputs is necessary, implying that I-S3 cell firing broadens the window for pyramidal cell disinhibition. Using in vivo virtual networks, we show that I-S3 cells contribute to a sharpening of OLM cell recruitment at θ frequencies. Furthermore, shifting the timing of I-S3 cell spiking due to external modulation shifts the timing of the OLM cell firing and thus disinhibitory windows. We propose a specialized contribution of I-S3 cells to create temporally precise coordination of modulation pathways.NEW & NOTEWORTHY How information is processed across different brain structures is an important question that relates to the different functions that the brain performs. Using modeling and theoretical analyses, we show that an inhibitory cell type that only inhibits other inhibitory cells can broaden the window for disinhibition of excitatory cells, manage input pathway switching, and modulate inhibitory cell spiking. This work contributes to the knowledge of how coordination between sensory and memory consolidation information can be attained.

Linking minimal and detailed models of CA1 microcircuits reveals how theta rhythms emerge and their frequencies controlled.

The wide variety of cell types and their biophysical complexities pose a challenge in our ability to understand oscillatory activities produced by cellular-based computational network models. This challenge stems from their high-dimensional and multiparametric natures. To overcome this, we implement a solution by linking minimal and detailed models of CA1 microcircuits that generate intrahippocampal (3-12 Hz) theta rhythms. We leverage insights from minimal models to guide explorations of more detailed models and obtain a cellular perspective of theta generation. Our findings distinguish the pyramidal cells as the theta rhythm initiators and reveal that their activity is regularized by the inhibitory cell populations, supporting a proposed hypothesis of an "inhibition-based tuning" mechanism. We find a strong correlation between input current to the pyramidal cells and the resulting local field potential theta frequency, indicating that intrinsic pyramidal cell properties underpin network frequency characteristics. This work provides a cellular-based foundation from which in vivo theta activities can be explored.

Synaptic Mechanisms Underlying the Network State-Dependent Recruitment of VIP-Expressing Interneurons in the CA1 Hippocampus.

Disinhibition is a widespread circuit mechanism for information selection and transfer. In the hippocampus, disinhibition of principal cells is provided by the interneuron-specific interneurons that express the vasoactive intestinal polypeptide (VIP-IS) and innervate selectively inhibitory interneurons. By combining optophysiological experiments with computational models, we determined the impact of synaptic inputs onto the network state-dependent recruitment of VIP-IS cells. We found that VIP-IS cells fire spikes in response to both the Schaffer collateral and the temporoammonic pathway activation. Moreover, by integrating their intrinsic and synaptic properties into computational models, we predicted recruitment of these cells between the rising phase and peak of theta oscillation and during ripples. Two-photon Ca2+-imaging in awake mice supported in part the theoretical predictions, revealing a significant speed modulation of VIP-IS cells and their preferential albeit delayed recruitment during theta-run epochs, with estimated firing at the rising phase and peak of the theta cycle. However, it also uncovered that VIP-IS cells are not activated during ripples. Thus, given the preferential theta-modulated firing of VIP-IS cells in awake hippocampus, we postulate that these cells may be important for information gating during spatial navigation and memory encoding.

Excitatory Inputs Determine Phase-Locking Strength and Spike-Timing of CA1 Stratum Oriens/Alveus Parvalbumin and Somatostatin Interneurons during Intrinsically Generated Hippocampal Theta Rhythm.

UNLABELLED: Theta oscillations are essential for learning and memory, and their generation requires GABAergic interneurons. To better understand how theta is generated, we explored how parvalbumin (PV) and somatostatin (SOM) interneurons in CA1 stratum oriens/alveus fire during hippocampal theta and investigated synaptic mechanisms underlying their behavior. Combining the use of transgenic mice to visually identify PV and SOM interneurons and the intact hippocampal preparation that can generate theta oscillations in vitro without cholinergic agonists, we performed simultaneous field and whole-cell recordings. We found that PV interneurons uniformly fire strongly phase-locked to theta, whereas SOM neurons are more heterogeneous with only a proportion of cells displaying tight phase-locking. Differences in phase-locking strength could be explained by disparity in excitatory inputs received; PV neurons received significantly larger EPSCs compared with SOM neurons, and the degree of phase-locking in SOM neurons was significantly correlated with the size of EPSCs. In contrast, IPSC amplitude did not differ between cell types. We determined that the local CA1 rhythm plays a more dominant role in driving CA1 interneuron firing than afferent inputs from the CA3. Last, we show that PV and strongly phase-locked SOM neurons fire near the peak of CA1 theta, under the tight control of excitatory inputs that arise at a specific phase of each theta cycle. These results reveal a fundamental mechanism of neuronal phase-locking and highlight an important role of excitation from the local network in governing firing behavior during rhythmic network states. SIGNIFICANCE STATEMENT: Rhythmic activity in the theta range (3-12 Hz) is important for proper functioning of the hippocampus, a brain area essential for learning and memory. To understand how theta rhythm is generated, we investigated how two types of inhibitory neurons, those that express parvalbumin and somatostatin, fire action potentials during theta in an in vitro preparation of the mouse hippocampus. We found that the amount of excitatory input they receive from the local network determines how closely their spikes follow the network theta rhythm. Our findings reveal an important role of local excitatory input in driving inhibitory neuron firing during rhythmic states and may have implications for diseases, such as epilepsy and Alzheimer's disease, which affect the hippocampus and related areas.

Building a mathematical model of the brain.

Automatic leveraging of information in a hippocampal neuron database to generate mathematical models should help foster interactions between experimental and computational neuroscientists.

Inhibitory networks of fast-spiking interneurons generate slow population activities due to excitatory fluctuations and network multistability.

Slow population activities (SPAs) exist in the brain and have frequencies below ~5 Hz. Despite SPAs being prominent in several cortical areas and serving many putative functions, their mechanisms are not well understood. We studied a specific type of in vitro GABAergic, inhibition-based SPA exhibited by C57BL/6 murine hippocampus. We used a multipronged approach consisting of experiment, simulation, and mathematical analyses to uncover mechanisms responsible for hippocampal SPAs. Our results show that hippocampal SPAs are an emergent phenomenon in which the "slowness" of the network is due to interactions between synaptic and cellular characteristics of individual fast-spiking, inhibitory interneurons. Our simulations quantify characteristics underlying hippocampal SPAs. In particular, for hippocampal SPAs to occur, we predict that individual fast-spiking interneurons should have frequency-current (f-I) curves that exhibit a suitably sized kink where the slope of the curve decreases more abruptly in the gamma frequency range with increasing current. We also predict that these interneurons should be well connected with one another. Our mathematical analyses show that the combination of synaptic and intrinsic conditions, as predicted by our simulations, promotes network multistability. Population slow timescales occur when excitatory fluctuations drive the network between different stable network firing states. Since many of the parameters we use are extracted from experiments and subsequent measurements of experimental f-I curves of fast-spiking interneurons exhibit characteristics as predicted, we propose that our network models capture a fundamental operating mechanism in biological hippocampal networks.

Reduced oriens-lacunosum/moleculare cell model identifies biophysical current balances for in vivo theta frequency spiking resonance.

Conductance-based models have played an important role in the development of modern neuroscience. These mathematical models are powerful "tools" that enable theoretical explorations in experimentally untenable situations, and can lead to the development of novel hypotheses and predictions. With advances in cell imaging and computational power, multi-compartment models with morphological accuracy are becoming common practice. However, as more biological details are added, they make extensive explorations and analyses more challenging largely due to their huge computational expense. Here, we focus on oriens-lacunosum/moleculare (OLM) cell models. OLM cells can contribute to functionally relevant theta rhythms in the hippocampus by virtue of their ability to express spiking resonance at theta frequencies, but what characteristics underlie this is far from clear. We converted a previously developed detailed multi-compartment OLM cell model into a reduced single compartment model that retained biophysical fidelity with its underlying ion currents. We showed that the reduced OLM cell model can capture complex output that includes spiking resonance in in vivo-like scenarios as previously obtained with the multi-compartment model. Using the reduced model, we were able to greatly expand our in vivo-like scenarios. Applying spike-triggered average analyses, we were able to to determine that it is a combination of hyperpolarization-activated cation and muscarinic type potassium currents that specifically allow OLM cells to exhibit spiking resonance at theta frequencies. Further, we developed a robust Kalman Filtering (KF) method to estimate parameters of the reduced model in real-time. We showed that it may be possible to directly estimate conductance parameters from experiments since this KF method can reliably extract parameter values from model voltage recordings. Overall, our work showcases how the contribution of cellular biophysical current details could be determined and assessed for spiking resonance. As well, our work shows that it may be possible to directly extract these parameters from current clamp voltage recordings.

Hyperpolarization-Activated Cation Channels Shape the Spiking Frequency Preference of Human Cortical Layer 5 Pyramidal Neurons.

Discerning the contribution of specific ionic currents to complex neuronal dynamics is a difficult, but important, task. This challenge is exacerbated in the human setting, although the widely characterized uniqueness of the human brain compared with preclinical models necessitates the direct study of human neurons. Neuronal spiking frequency preference is of particular interest given its role in rhythm generation and signal transmission in cortical circuits. Here, we combine the frequency-dependent gain (FDG), a measure of spiking frequency preference, and novel in silico analyses to dissect the contributions of individual ionic currents to the suprathreshold features of human layer 5 (L5) neurons captured by the FDG. We confirm that a contemporary model of such a neuron, primarily constrained to capture subthreshold activity driven by the hyperpolarization-activated cyclic nucleotide gated (h-) current, replicates key features of the in vitro FDG both with and without h-current activity. With the model confirmed as a viable approximation of the biophysical features of interest, we applied new analysis techniques to quantify the activity of each modeled ionic current in the moments before spiking, revealing unique dynamics of the h-current. These findings motivated patch-clamp recordings in analogous rodent neurons to characterize their FDG, which confirmed that a biophysically detailed model of these neurons captures key interspecies differences in the FDG. These differences are correlated with distinct contributions of the h-current to neuronal activity. Together, this interdisciplinary and multispecies study provides new insights directly relating the dynamics of the h-current to suprathreshold spiking frequency preference in human L5 neurons.

Network bursting using experimentally constrained single compartment CA3 hippocampal neuron models with adaptation.

The hippocampus is a brain structure critical for memory functioning. Its network dynamics include several patterns such as sharp waves that are generated in the CA3 region. To understand how population outputs are generated, models need to consider aspects of network size, cellular and synaptic characteristics and context, which are necessarily 'balanced' in appropriate ways to produce particular outputs. Thick slice hippocampal preparations spontaneously produce sharp waves that are initiated in CA3 regions and depend on the right balance of glutamatergic activities. As a step toward developing network models that can explain important balances in the generation of hippocampal output, we develop models of CA3 pyramidal cells. Our models are single compartment in nature, use an Izhikevich-type structure and involve parameter values that are specifically designed to encompass CA3 intrinsic properties. Importantly, they incorporate spike frequency adaptation characteristics that are directly comparable to those measured experimentally. Excitatory networks using these model cells are able to produce bursting suggesting that the amount of spike frequency adaptation expressed in the biological cells is an essential contributor to network bursting, and as such, may be important for sharp wave generation. The network bursting mechanism is numerically dissected showing the critical balance between adaptation and excitatory drive. The compact nature of our models allows large network simulations to be efficiently computed. This, together with the linkage of our models to cellular characteristics, will allow us to develop an understanding of population output of CA3 hippocampus with direct biological comparisons.

Tackling recalcitrant Pseudomonas aeruginosa infections in critical illness via anti-virulence monotherapy.

Intestinal barrier derangement allows intestinal bacteria and their products to translocate to the systemic circulation. Pseudomonas aeruginosa (PA) superimposed infection in critically ill patients increases gut permeability and leads to gut-driven sepsis. PA infections are challenging due to multi-drug resistance (MDR), biofilms, and/or antibiotic tolerance. Inhibition of the quorum-sensing transcriptional regulator MvfR(PqsR) is a desirable anti-PA anti-virulence strategy as MvfR controls multiple acute and chronic virulence functions. Here we show that MvfR promotes intestinal permeability and report potent anti-MvfR compounds, the N-Aryl Malonamides (NAMs), resulting from extensive structure-activity-relationship studies and thorough assessment of the inhibition of MvfR-controlled virulence functions. This class of anti-virulence non-native ligand-based agents has a half-maximal inhibitory concentration in the nanomolar range and strong target engagement. Using a NAM lead in monotherapy protects murine intestinal barrier function, abolishes MvfR-regulated small molecules, ameliorates bacterial dissemination, and lowers inflammatory cytokines. This study demonstrates the importance of MvfR in PA-driven intestinal permeability. It underscores the utility of anti-MvfR agents in maintaining gut mucosal integrity, which should be part of any successful strategy to prevent/treat PA infections and associated gut-derived sepsis in critical illness settings. NAMs provide for the development of crucial preventive/therapeutic monotherapy options against untreatable MDR PA infections.

Corrigendum: Inhibitory Network Bistability Explains Increased Interneuronal Activity Prior to Seizure Onset.

[This corrects the article DOI: 10.3389/fncir.2019.00081.].

A Hypothesis for Theta Rhythm Frequency Control in CA1 Microcircuits.

Computational models of neural circuits with varying levels of biophysical detail have been generated in pursuit of an underlying mechanism explaining the ubiquitous hippocampal theta rhythm. However, within the theta rhythm are at least two types with distinct frequencies associated with different behavioral states, an aspect that must be considered in pursuit of these mechanistic explanations. Here, using our previously developed excitatory-inhibitory network models that generate theta rhythms, we investigate the robustness of theta generation to intrinsic neuronal variability by building a database of heterogeneous excitatory cells and implementing them in our microcircuit model. We specifically investigate the impact of three key "building block" features of the excitatory cell model that underlie our model design: these cells' rheobase, their capacity for post-inhibitory rebound, and their spike-frequency adaptation. We show that theta rhythms at various frequencies can arise dependent upon the combination of these building block features, and we find that the speed of these oscillations are dependent upon the excitatory cells' response to inhibitory drive, as encapsulated by their phase response curves. Taken together, these findings support a hypothesis for theta frequency control that includes two aspects: (i) an internal mechanism that stems from the building block features of excitatory cell dynamics; (ii) an external mechanism that we describe as "inhibition-based tuning" of excitatory cell firing. We propose that these mechanisms control theta rhythm frequencies and underlie their robustness.

Characterization of voltage-gated K+ currents contributing to subthreshold membrane potential oscillations in hippocampal CA1 interneurons.

CA1 inhibitory interneurons at the stratum lacunosum-moleculare and radiatum junction (LM/RAD-INs) display subthreshold membrane potential oscillations (MPOs) involving voltage-dependent Na(+) and A-type K(+) currents. LM/RAD-INs also express other voltage-gated K(+) currents, although their properties and role in MPOs remain unclear. Here, we characterized these voltage-gated K(+) currents and investigated their role in MPOs. Using outside-out patch recordings from LM/RAD-IN somata, we distinguished four voltage-gated K(+) currents based on their pharmacology and activation/inactivation properties: a fast delayed rectifier current (I(Kfast)), a slow delayed rectifier current (I(Kslow)), a rapidly inactivating A-type current (I(A)), and a slowly inactivating current (I(D)). Their relative contribution to the total K(+) current was I(A) > I(Kfast) > I(Kslow) = I(D). The presence of I(D) and the relative contributions of K(+) currents in LM/RAD-INs are different from those of other CA1 interneurons, suggesting the presence of differential complement of K(+) currents in subgroups of interneurons. We next determined whether these K(+) currents were sufficient for MPO generation using a single-compartment model of LM/RAD-INs. The model captured the subthreshold voltage dependence of MPOs. Moreover, all K(+) currents were active at subthreshold potentials but I(D), I(A), and the persistent sodium current (I(NaP)) were most active near threshold. Using impedance analysis, we found that I(A) and I(NaP) contribute to MPO generation by modulating peak spectral frequency during MPOs and governing the voltage range over which MPOs occur. Our findings uncover a differential expression of a complement of K(+) channels that underlies intrinsic rhythmic activity in inhibitory interneurons.

Spatial coherence and stationarity of local field potentials in an isolated whole hippocampal preparation in vitro.

Local field potential (LFP) multielectrode recordings of spontaneous rhythms in an isolated whole hippocampal preparation are characterized with respect to their spatial variability within the hippocampus, and their frequency properties. Using simulated data, we categorize potential relationships between frequency variation over time in LFP recordings and spatial variability between electrodes. We then use data recorded from the intact preparation to distinguish between our theoretical categories. We find that the LFP recordings have a close to spatially invariant frequency distribution (not phase) across the hippocampus, and differ in frequency only in a component that may be seen as physiological noise. From these facts, we conclude that the isolated hippocampal LFP recordings represent a single signal and may be regarded as a unitary circuitry. We additionally examine phase differences across our recording sites. We use our characterization of the hippocampal isolate's properties to predict its spatial coherence in response to high frequency stimulation. We find that there is a finely tuned inverse relationship between temporal variability in the hippocampal isolate's LFP recordings and their spatial coherence.

Common Principles in Functional Organization of VIP/Calretinin Cell-Driven Disinhibitory Circuits Across Cortical Areas.

In the brain, there is a vast diversity of different structures, circuitries, cell types, and cellular genetic expression profiles. While this large diversity can often occlude a clear understanding of how the brain works, careful analyses of analogous studies performed across different brain areas can hint at commonalities in neuronal organization. This in turn can yield a fundamental understanding of necessary circuitry components that are crucial for how information is processed across the brain. In this review, we outline recent in vivo and in vitro studies that have been performed in different cortical areas to characterize the vasoactive intestinal polypeptide (VIP)- and/or calretinin (CR)-expressing cells that specialize in inhibiting GABAergic interneurons. In doing so, we make the case that, across cortical structures, interneuron-specific cells commonly specialize in the synaptic disinhibition of excitatory neurons, which can ungate the integration and plasticity of external inputs onto excitatory neurons. In line with this, activation of interneuron- specific cells enhances animal performance across a variety of behavioral tasks that involve learning, memory formation, and sensory discrimination, and may represent a key target for therapeutic interventions under different pathological conditions. As such, interneuron-specific cells across different cortical structures are an essential network component for information processing and normal brain function.

Computationally going where experiments cannot: a dynamical assessment of dendritic ion channel currents during in vivo-like states.

Background: Despite technological advances, how specific cell types are involved in brain function remains shrouded in mystery. Further, little is known about the contribution of different ion channel currents to cell excitability across different neuronal subtypes and their dendritic compartments in vivo. The picture that we do have is largely based on somatic recordings performed in vitro. Uncovering dendritic ion channel current contributions in neuron subtypes that represent a minority of the neuronal population is not currently a feasible task using purely experimental means. Methods: We employ two morphologically-detailed multi-compartment models of a specific type of inhibitory interneuron, the oriens lacunosum moleculare (OLM) cell. The OLM cell is a well-studied cell type in CA1 hippocampus that is important in gating sensory and contextual information. We create in vivo-like states for these cellular models by including levels of synaptic bombardment that would occur in vivo. Using visualization tools and analyses we assess the ion channel current contribution profile across the different somatic and dendritic compartments of the models. Results: We identify changes in dendritic excitability, ion channel current contributions and co-activation patterns between in vitro and in vivo-like states. Primarily, we find that the relative timing between ion channel currents are mostly invariant between states, but exhibit changes in magnitudes and decreased propagation across dendritic compartments. We also find enhanced dendritic hyperpolarization-activated cyclic nucleotide-gated channel (h-channel) activation during in vivo-like states, which suggests that dendritically located h-channels are functionally important in altering signal propagation in the behaving animal. Conclusions: Overall, we have demonstrated, using computational modelling, the dynamical changes that can occur to ion channel mechanisms governing neuronal spiking in vitro and in vivo. In particular, we have shown that the magnitudes of some ion channel current contributions are differentially altered during in vivo-like states relative to in vitro.